SwitchLight: Co-Design of Physics-Driven Architecture and Pre-Training Framework for Human Portrait Relighting
Abstract
We introduce a co-designed approach for human portrait relighting that combines a physics-guided architecture with a pre-training framework. Drawing on the Cook-Torrance reflectance model we have meticulously configured the architecture design to precisely simulate light-surface interactions. Furthermore to overcome the limitation of scarce high-quality lightstage data we have developed a self-supervised pre-training strategy. This novel combination of accurate physical modeling and expanded training dataset establishes a new benchmark in relighting realism.
Cite
Text
Kim et al. "SwitchLight: Co-Design of Physics-Driven Architecture and Pre-Training Framework for Human Portrait Relighting." Conference on Computer Vision and Pattern Recognition, 2024. doi:10.1109/CVPR52733.2024.02371Markdown
[Kim et al. "SwitchLight: Co-Design of Physics-Driven Architecture and Pre-Training Framework for Human Portrait Relighting." Conference on Computer Vision and Pattern Recognition, 2024.](https://mlanthology.org/cvpr/2024/kim2024cvpr-switchlight/) doi:10.1109/CVPR52733.2024.02371BibTeX
@inproceedings{kim2024cvpr-switchlight,
title = {{SwitchLight: Co-Design of Physics-Driven Architecture and Pre-Training Framework for Human Portrait Relighting}},
author = {Kim, Hoon and Jang, Minje and Yoon, Wonjun and Lee, Jisoo and Na, Donghyun and Woo, Sanghyun},
booktitle = {Conference on Computer Vision and Pattern Recognition},
year = {2024},
pages = {25096-25106},
doi = {10.1109/CVPR52733.2024.02371},
url = {https://mlanthology.org/cvpr/2024/kim2024cvpr-switchlight/}
}